DNS 1-3 Quantification of Resolution: Difference between revisions
Line 10: | Line 10: | ||
==Solution verification== | ==Solution verification== | ||
The 10000 Reynolds number diffuser has been run and compared with the results with previous DNS data performed by [[lib:DNS_1-3_description#1|Ohlsson ''et al.'' (2010)]]. A P2 mesh of 250 Million of DoF. The resolution is fine enough to be at DNS level. An animation of the fluctuations of this flow can be seen in [http://kbwiki-data.s3-eu-west-2.amazonaws.com/DNS-1/3/fluct2. | The 10000 Reynolds number diffuser has been run and compared with the results with previous DNS data performed by [[lib:DNS_1-3_description#1|Ohlsson ''et al.'' (2010)]]. A P2 mesh of 250 Million of DoF. The resolution is fine enough to be at DNS level. An animation of the fluctuations of this flow can be seen in [http://kbwiki-data.s3-eu-west-2.amazonaws.com/DNS-1/3/fluct2.mp4 fluctuations]. | ||
<div id="figure2"></div> | <div id="figure2"></div> |
Revision as of 09:51, 16 February 2021
Quantification of resolution
Mesh resolution
A mesh resulting in approximately 250 million degrees of freedom (DoF). With a stretched grid, the maximum grid resolution in the duct centre is reported at , , . Correspondingly, the wall resolution (in terms of the first grid point) is reported as , in the spanwise and normal directions, respectively, is used.
Solution verification
The 10000 Reynolds number diffuser has been run and compared with the results with previous DNS data performed by Ohlsson et al. (2010). A P2 mesh of 250 Million of DoF. The resolution is fine enough to be at DNS level. An animation of the fluctuations of this flow can be seen in fluctuations.
Figure 2: Stanford double diffuser, Re=10000, preliminary validation. Average streamwise velocity, Ohlsson et al. (2010) vs preliminary data obtained with Alya. |
Figure 3: Stanford double diffuser, Re=10000, preliminary validation. Average streamwise velocity fluctuations, Ohlsson et al. (2010) vs preliminary data obtained with Alya. |
In Fig. 2 and Fig. 3 results are presented with reference data from Ohlsson et al. (2010). Fair agreement is observed between both calculations, proving that the approach to be used in the present proposal is an optimal strategy. Turbulent inlet was generated in a precursor domain using a long enough duct with a roughness element to trigger the laminar to turbulent transition.
Contributed by: Oriol Lehmkuhl, Arnau Miro — Barcelona Supercomputing Center (BSC)
© copyright ERCOFTAC 2024